Morphology and Epithelial Ion Transport of the Alkaline Gland in the Atlantic Stingray (Dasyatis sabina)

12 Pages·1999·5.6 MB·English

by  G M Grabowski

Reference: Biol. Bull 197: 82-93. (August 1999]
Morphology and Epithelial Ion Transport of the
Alkaline Gland in the Atlantic Stingray
(Dasyatis sabina)
GREGORY M. GRABOWSKI.1 JOHN G. BLACKBURN,2 AND ERIC R. LACY3'4
Department ofBiology, University ofDetroit Mercy, 4001 W. McNichols, P.O. Box 19900, Detroit,
Michigan 48219; 2 Department ofPhysiology, 3 Department ofCell Biologv and Anatomy,
Medical University ofSouth Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425;
and 4 Marine Biomedicine and Environmental Sciences, Medical University ofSouth Carolina,
221 Fort Johnson Road, Charleston, South Carolina 29412
Abstract. The alkaline glands are two fluid-filledsacs that the tissue caused the Isc to decrease >100%. Serosal expo-
lie on the ventral, posterior surface ofeach kidney in skates sure to ouabain (Na-K, ATPase inhibitor) decreased Isc
and rays. In this study, the morphology, transepithelial ion 48%, whereas amiloride (sodium ion channel blocker) and
transport, fluid constituents, and histochemistry ofthe alka- acetazolamide (carbonic anhydrase inhibitor) had no statis-
line glands of the Atlantic stingray, Dasyatis sabina, were tically significant effect on Isc oralkalinization rates. Taken
investigated. The duct from each gland joined the corre- together the results suggest the presence ofapical epithelial
sponding vas deferens and the resulting two common ducts bicarbonate exchangers that are chloride or sodium depen-
emptied into the cloaca. Dark burgundy, aqueous fluid (pH dent, basal sodium and HCO^ transport, and an Isc that is
8.0-8.2) was secreted into the sacs by a simple columnar not totally dependent on Na+-K+ ATPase.
epithelium with extensive rough endoplasmic reticulum and
large secondary lysosomes containing lipofuscin and mem- Introduction
brane fragments. Zonulae occludentes were deep (22
fibrils), reflecting an electrically tight epithelium (732 Early anatomical studies of the male skate and stingray
ohms/cm2). Carbonic anhydrase activity was localized his- urogenital system reported a pair of blind-ended sacs, each
tochemically within the intercellular spaces and less in- of which opened into the cloaca. These structures were
tensely in the mid-basal cytoplasm. described initially as urinary bladders or sperm storage sacs
In vitro electrophysiology showed that baseline short- (Borcea, 1906; Daniel, 1934), but the only evidence to
circuit current (Isc, 29.1 /AA/cm2) was reduced 67.0% after support this functional nomenclature is the proximity ofthe
Cl~ removal from the medium. Cl removal also com- sacs' openings to those of the ureters and vas deferens
pletely abolished luminal alkalinization (baseline 4.5 0.7 within the cloaca.
/LtEq of acid/cnr/h). Luminal exposure to the chloride- The sacs secrete and store a watery fluid of high pH
bicarbonate exchange inhibitor, DIDS, reduced Isc by 38%. (8.0-9.2), thus their name, alkaline gland (Maren et al.,
Simultaneous administration of DIDS and bumetanide 1963). Onthebasisofthe high pHofthe fluid, Smith (1929)
(Na+/K+/Cl~ cotransport inhibitor) to the serosal side of speculated that it neutralized the potentially deleterious
effects ofacidic urine in the cloaca on the extruded sperm.
As yet, however, nostudies on the physiological function of
Received 12 April 1999; accepted 14 June 1999. the alkaline gland have been published.
Send correspondence to Eric R. Lacy. Marine Biomedicine and Envi- A few reports, from various skate species (little skate,
ronmental Sciences, Medical UniversityofSouthCarolina, 221 FortJohn-
son Road. Charleston. SC 29412. Raja erinacea; barndoor skate, R. stabuliforis; big skate, R.
A portion of this work was presented in abstract form (The FASEB ocellata), have described the gland's morphology and epi-
Journal, Part I, #3024, 1992). thelial transport physiology (H.W. Smith. 1929; Maren et
82
STINGRAY ALKALINE GLAND
83
al.. 1963; Masur. 1984; P. L. Smith, 1981, 1985). These ester, 0.5 g/1, Sigma Chemical Co.) and double pithed. The
morphological accounts show that the gland lumen has body cavity was opened by a ventral midline incision; the
mucosal "villar projections" lined by a simple columnar alkaline gland fluid was aspirated with a 25-gauge needle
epithelium (Maren et al.. 1963; Masur, 1984). The mucosa and saved at 4C forfurtheranalysis; the alkaline gland was
dgieennetraotfesOHanidonmsaianntdaian5s0-afohludndgrreaddi-efnotldofcConOce2nftrraotmiopnlagsrmaa- irmeemnotvs.ed for use in morphology orelectrophysiology exper-
to gland lumen; these are some of the steepest alkaline
gradients across any epithelium in nature (Maren el al., Light and electron microscopy
1963). Given the unique epithelial transport properties of
the alkaline gland, physiologic studies have focused on the Fixative (2.5% paraformaldehyde, 5.0% glutaraldehyde,
mechanisms of fluid and bicarbonate secretion (Maren et and 0.25% picric acid; Ito and Karnovsky. 1968) was in-
al., 1963; Smith, 1981, 1985). jected intoboth sacs ofthe gland immediately afterthe fluid
Chloride and bicarbonate are the two main anions con- was removed. After 1 h the puboischiac bar was severed,
stituting alkaline gland fluid in the skate. In vitro experi- and the alkaline gland was freed from surrounding tissue
ments indicate that chloride secretion accounts for most, if with fine forceps. Each gland was excised at its junction
not all, ofthe short-circuit current (Isc) (Maren etal.. 1963; with the cloacal wall and placed in the same fixative for
Smith. 1981. 1985). These results led to speculation that 24h. Thetissue wasthen rinsed, trimmMed into 1-mnr pieces
chloride-dependent bicarbonate transport might be involved with a razor blade, and stored in 0.1 sodium cacodylate
buffer.
in fluid alkalinization. Although definitive evidence was
lacking, secreted chloride was believed to recirculate into Alkaline gland fluid was centrifuged at 200 X g for 10
theepithelial cell by way ofaCr/HCOJ exchangerlocated min. The pellet was fixed for 4 h in the same fixative
at the apical plasma membrane (Maren et al., 1963; Smith, injected intothe gland sacs (Ito and Karnovsky, 1968). Both
1981. 1985). Carbonic anhydrase, an enzyme associated pellet and pieces of fixed glMand were then postfixed (1.0%
with many bicarbonate-secreting tissues, was identified bio- osmium tetraoxide in 0.1 sodium cacodylate buffer),
chemically in the alkaline gland of some but not all skate dehydrated in graded ethanols, and embedded in Epon-
species studied (Maren et al., 1963). The concentration of Araldite. Sections were cut, stained (semithin sections
carbonic anhydrase in the tissue was correlated with the pH stained with alkalinized toluidine blue and ultrathin sections
of the alkaline gland fluid produced (Maren et al., 1963), with uranyl acetate and lead citrate), and examined using a
suggesting that this enzyme has a role in bicarbonate secre- light microscope or a JEOL 1200 EX electron microscope.
tion for some skate species. Additional gland tissue, fixed as described above but in
The present study uses transmission and scanning elec- aldehydes only, was cryoprotected in graded concentrations
tron microscopy and freeze fracture to elucidate the ultra- of glycerols to a final concentration of 30% glycerol for
structural organization of the alkaline gland in a stingray freeze fracture. The tissue was then frozen rapidly in liquid
species, Dasyatis sabina, the Atlantic stingray. The pres- propaMne, followed by fracturing and replication in a Balzer
ence and distribution ofcarbonic anhydrase activity, nerve 360 device (Balzers, Fiirstentum Liechtenstein). Replicas
fibers, andlipofusionwereidentifiedhistochemically. These were supported on 200-mesh copper grids and examined
results are correlated with in vitro electrophysiological data with the transmission electron microscope.
and rates of fluid alkalinization. Some of the regulatory Aldehyde-fixedtissue was alsoused forscanningelectron
mechanism of ion transport were probed with various met- microscopyM. It was first postfixed in 1.0% osmium tetraox-
abolic inhibitors. The composition of the fluid removed ide in 0.1 sodium cacodylate buffer, followed by dehy-
from the alkaline glands was analyzed. dration in graded ethanols, and then critical point dried
using a Sorvall critical point dryer (Newtown, CT). Tissue
was coated with gold/palladium for 3 min at 2.5 kV and 20
Materials and Methods mA
using an E5100 sputter coating unit (Polaron Instru-
Sexually mature male Atlantic stingrays (Dasvatis sa- ments, Doylestown, PA) and examined with a JEOL 35C
bina, wing span ~45 cm) purchased from Gulf Specimens scanning electron microscope.
Inc. (Panacea, FL) or captured along the coast of South
Carolina were allowed to acclimate in a 16,000-1 holding Lipofuscin staining
tank for at least 5 days prior to experimentation. Water in Alkaline gland tissue and paniculate matter from gland
the holding tank was drawn from Charleston (South Caro- fluid offour stingrays were stained for lipofuscins using the
lina) Harbor (650-850 mosm/1) and maintained at room Long Ziehl-Neelsen technique (Bancroft and Cook, 1984).
temperature. Stingrays were fed shrimp twice a week and The pellet, as described above, and gland tissue were fixed
kept on a 12-h light/dark cycle. After acclimation, animals in Bouin's solution for 2 h, followed by dehydration in
were anesthetized with MS222 (3-aminobenzoic acid ethyl graded ethanols, clearing in xylene, and embedding in par-
84 G. M. GRABOWSKI ET AL
aftin. Five-micrometer-thick sections were deparaffinized in Morplioinctric analysis
xylene taken stepwise to water and stained in filtered carbol Ratios of basal cells to columnar cells were determined
fuchsin for 1-3 h at 56C. After staining, sections were from counts made of cross-sectioned glands at the light
washed in water, differentiated in 1% acid-alcohol, and microscopic level (epoxy resin sections, 50X). The size and
counterstained in aqueous methylene blue. Slideswere then distributionofintramembranous particlesobserved in freeze
rinsed in water, dehydrated, cleared in xylene. and mounted fracture replicas were measured on electron micrographs
on glass slides. Lipofuscin appeared bright magenta, and using a scale magnification loupe (Baxter. Atlanta, GA).
nuclei stained blue against a pale magenta background. The luminal surface area of columnar epithelial cells was
estimated by measuring the cell diameters of luminal
Silver staining ofneural tissue plasma membranes from scanning electron micrographs.
Nerve fibers in alkaline glands were localized using the Constituents ofalkaline gland fluid
silver precipitate method of Sevier and Munger (1965).
Five-micrometer-thick paraffin sections of Bouin's fixed Fluid from the alkaline glands of five stingrays was
tissue were incubated in 20% silver nitrate for 15 mm, pooled, cooled to 5C, and centrifuged as described above.
washed with distilled water, and developed in ammoniacal The supernatant was then frozen by placing the tube in dry
silver (10% silver nitrate precipitated with 28%-30% am- ice and shipped overnight to Mayo Medical Laboratories
monium hydroxide, plus 2% formalin). After a 2-min rinse (Rochester. MN) for analysis of its composition.
in 5% sodium thiosulfate. slides were washed in distilled
water, dehydrated, cleared in xylene, and mounted. Electrophysiology
Each sac of the alkaline gland was freed in situ from
Localization ofcarbonic anhydrase activity (CAM) suiTOunding connective tissue, excised, and placed in a petri
dish ofoxygenated elasmobranch Ringer(NaCl. 280.0 mM;
malAdlkeahlyidnee,g2l.a5nd%sgwleutraeraflidxeedhyidne,aasnoldut0i.o4n%ofC2a.C0K%ipnar0a.1forM- K1C0.10.m5M.0;muMre;a,M3g50C.0Um3M.3;mdeMx;troCsaeC,l52..03m.8M;mM8;00NmaOHsmC/O1;v
sodium cacodylate buffer for localization ofcarbonic anhy- pH 6.9). The Ringer was gassed with 95% O:/5% CO2,
drase activity (CAH) using the Hansson's technique (Hans- unless otherwise noted, and used at room temperature.
son, 1967: Maren. 1980b: Sugai and Ito, 1980; Lacy, Each sac was mounted between two halves ofan Ussing
1983b). Fixed tissue was frozen in 8% sucrose and sec- chamber (4-mm diameter). Each half of the chamber was
tioned at 10 M111 on an IEC CTF cryostat (International connected to a 20-ml circulation reservoir (Medical Re-
Emqeudiipumment(1.C8o6mpmanMy)mC.MoSSecOt4i.on5s5.w9ermeMfloaHt2eSdOo4n, H3a.n7s3somn'Ms s(Iesacr)chanAdpptararnasteupsi,thCellieaalrwpaotteern,tiFalL)d.ifTfheeresnhcoert(-PcDir)cuwietrceurmreean-t
KH,PO4, and 158 NaHCO,) for 1-5 min. Sections sured using a voltage-current clamp (Physiological Instru-
were rinsed by floating on Sorensen's phosphate buffer (pH ments. San Diego, CA). Before tissue was mounted in the
8.0) for 1 min and then transferred onto 2% ammonium Ussing chamber, electrode polarization and fluid resistance
sulfide for 1-2 min. This was followed by rinsing sections was compensated with the VCC600 voltage-current clamp.
on Sorensen's phosphate buffer at pH 5.0 and then mount- Calomel electrodes (Fisher, Atlanta, GA) placed in a satu-
ing them in heated glycerin jelly on glass slides for obser- rated KC1 solution were connected to the Ussing chamber
v1a9t8i3obn).wiAtfhtearltighhetsmeicctrioonsscowpeere(SuignaciubaantdedItoo,n 129%80;amLamcoy-, vsiuaresatlhtebPrDi.dgPelsat(i4n%umagealercitnroedleass(mFoibshrearn.cAhtlRainntgae,rG)At)o mweear-e
pnrieucmipistulaftiedei,ndliocwat-ipvHe boufffCeArsHwearcteivuisteydftroopmredveegnrtadtihneg.blFacokr pPlDaceadnddirIescctlwyeirnetodtihseplUasyseidngonchaamSboelrtetco m1e2a4s2ursetrIispc.chTahret
electron microscopy, sections were postfixed in 1.0% os- recorder (Soltec Corp.. Sun Valley, CA). Transepithelial
mium tetraoxide in Sorensen's phosphate buffer (pH 5.0) resistance was calculated using the open-circuit PD, and the
for 30-45 min, stained en bloc with 1.0% uranyl acetate in closed-circuit Isc of the mounted tissue. All readings were
maleate buffer (pH 5.2). dehydrated in graded ethanols, and in reference to the luminal medium.
embedded flat in epoxy resin. Ultrathin sections were
stained and examined as described above. Transport inhibitors
Acetazolamide 10~5 and 10~6 M) in Hansson's medium
was used to inhib(it CAH, thereby serving as a negative Once baseline electrophysiological parameters were es-
control. Forevaluation ofnonspecific activity, sections were tablished, the percent change ofIsc was calculated after the
incubated either in ammonium sulfide without prior incu- tissue was exposed to the following transport inhibitors:
bation on Hansson's medium, or on bicarbonate-free Hans- ouabain. Na"/K+ ATPase inhibitor ( 10~4 M. serosal): bu-
son's medium. metanide, Na+/K+/Cl cotransport inhibitor (10 3 M. se-
STINGRAY ALKALINE GLAND 85
rosal); DIDS (4,4'-diisothiocyanatostilbene-2,2'-disulfonic ofthe kidney and lateral to the vas deferens. In the animals
acid, CT/HCO^ exchange inhibitor (10"' M, luminal); we examined, the glands were retroperitoneal and symmet-
amiloride, sodium channel inhibitor, (10 3 M. luminal and rically aligned along the vertebral column. They were easily
serosal); acetazolamide, carbonic anhydrase inhibitor (10~5 distinguished from surrounding tissue by their deep bur-
M, luminal). Chloride was substituted in the medium with gundy color. Each sac ofthe gland held a maximum of4-5
isomolar concentrations of gluconate. All reagents were ml of fluid. The mediocaudal portion of each gland nar-
purchased from Sigma Chemical Co.. St. Louis, MO. rowed to a single duct, which joined the respective sperm
duct (vas deferens) on the same side of the animal. The
Alkalinization rates resultant common duct for sperm and alkaline gland fluid
The alkalinization rate of the luminal medium was mea- was about 3-mm long and pierced the body wall to open on
sured using the pH stat technique on glands mounted in the the crest of the urinary papilla in the cloaca.
Ussing chamber. Unbuffered (bicarbonate-free) Ringer
bathing the luminal side ofthe gland was gassed with 100% Microscopv
eomroxeuydngteeinlngad.sumrToibhnregansecexhrpoesRraiiln-mgbeeanrtt,hsinggaasnmsdeedd3i0wuimtmhicno9n5spri%isotroedxtyoogfetnbiu/sfs5fu%-e linTehdebymuacossimaploef tchoelualmknaalrineepgiltahenldiuwmas(Fhiiggs.hly1,f2o)l.deAdraincdh
carbon dioxide. The rate of fluid alkalinization (/u,Eq Mof cmaepmibllraarnye.neWtiwtohriknleaaychimfmoelddiawteerleyabnenaretaetrhiolteheanbdasveemneunlte
acid/cnr/h) was then determined via titration using 0.01
sulfuric acid. The pH of the luminal medium was main- andTwdoenspoeputrlaacttisonosfonferevpeithfeilbiearls c(eFlilgss.we1r-e3).distinguished on
tained at pH 5.5 for at least six consecutive intervals of 5
(mmMiainncrepoaecHlhe.mceTttrhoeedrep,(HOLmwoeansgdaomno4dn0ei,rtrFoyur,leldeNruHtso)inn,cgoCanAn)pe.HctmeidcrtooelaecBtercokd-e dt(Fhiiefgf.ebra1esd)i.sTsihogfenitlfheiecniagrntthalpyoifciantlhtehmeleotmnwbgoraaxpinospeuolfeaxttphioeosnauspri(ecPatlo<cetl0hl.e0s5lu,rufnmaec=ne
Two experiments were performed to determine the pres- a1p4i1c)a;licnellonseur(fa8c4e.4w%asof7.t2he to0t.a1l4c^elml;s)itnhteheloontghearxi(s15o.f6%t)h,e
ence of either chloride-dependent or sodium-dependent bi-
ctearrbeoancahtemtarnainpsuploratt.ioAnl.kalBiansiezlaitnieonvraaltueesswweerreememaasdueredfraofm- t/uh,em).lonAgllaxciesllswatshaatbocuotnttawcitceed tthhaet lleunmgetnh (h1a4d.92the s0a.m4e9
tissues bathed on both sides with elasmobranch Ringer. The muletnraasltrmuectmubrarlanoergaarneiaz.ation, despite the difference in lu-
medium was changed on the luminal and serosal sides to
iso-osmotic elasmobranch Ringer free of chloride or so- Columnar epithelial cells had a prominent, basally lo-
cated pleomorphic nucleus and exceptionally large and
dwiausm.adIdnedthbeacfkirsttoetxhpeerliummeinnta,l cshidleo;riidne-tchoentsaeicnoinndg eRxipnegreir- abundant secondary lysosomes (Figs. 1, 2). The secondary
ment, sodium-containing Ringer was added back to the lysosomes stained positively forlipofuscin (data not shown)
serosal side. Afteranew alkalinization rate was established, and were dark green-brown in unstained sections. (Fig. 3).
The smooth-surfaced endoplasmic reticulum was evenly
a bicarbonate transport inhibitor, SITS (4-acetamido-4'-
isothiocyanatostilbene-2,2'-disulfonic acid, 10~3 M), was distributed throughout the cytoplasm. Mitochondria bearing
added to the luminal medium in both experiments. lamellar cristae were located in the upper two-thirds of the
bmuifnTfehederianbfgutfefrceaerpaianccghitceyaxppaoecfriitemyaencothf,tmuheseidvniagurtimhoeuwspaHRsinsttgaheternmsewtshaubostdr.daectTtehered- pcmrereeleglmsi,beoranntandni(enFG.itoghl.eSgoiG2mo)c.leogmioMpfraleentgxhyieeossnemwaevenmerdsbeiracdaaljbneausecn-ewdnbaetonrutteonidtsnheeetvhnaeepsifipcuceaslrlieisnnpgulcwawlesiermtaeahr
from the alkalinization rate derived under the experimental larger vesicles as well as with the apical plasma membrane
conditions.
(Fig. 2).
Basal cells were also present in the lower third of the
Statistical analyses
epithelium (Fig. 4) in a ratio of about 1 basal cell to 20
Statistical significance was evaluated using a two-tailed- columnar cells. These cells, which ranged from 1.4 to 2.6
paired t test, with the level of significance set at P < 0.05. /j,m indiameter, were not highly interdigitated with adjacent
columnar cells and were not observed in contact with an-
Results other basal cell. The cytoplasm of basal cells surrounded a
Gross anatom\ proportionately large nucleus and contained only a few
organelles, which were limited to the endoplasmic reticu-
The alkaline gland of the Atlantic stingray, Dasyatis lum, and small vesicles containing material of various de-
sabina, consists of a pair of blind-ended, bladder-like sacs grees of electron density.
located within the pelvic girdle ventral to the posterior pole The apical surface of the columnar cells was elaborated
86 G. M. GRABOWSKI ET AL
Figure 1. Figure 3.
.
STINGRAY ALKALINE GLAND 87
into microplicae (Figs. 1, 2). The basolateral plasma mem- phosphate buffer, and showed no positive staining (data not
CAH
brane was relatively straight nearest the lumen, but closerto shown). Complete inhibition oMf occurred at acetazol-
thebasal laminaitwas interdigitatedwith itselfandadjacent amide concentrationsof 10~5 in Hansson's medium (data
cells (Fig. 2). Freeze fracture of the lateral plasma mem- not shown). Lower concentrations of acetazolamide (10~6
brane revealed some areas consisting only of large in- M) failed to inhibit CAH activity for incubation periods
tramembranous particles (99 A 0.1, n = 52) (Fig. 5) longer than 2 min.
loosely arranged as single particles or in groups ofup to 20
particles. Outside these areas was a mixture of large and Analysis ofalkaline glandfluid (AGF)
small intramembranous particles. No rod-shaped particles
wjmaeemrm,ebrnoabn=see.r1v9Tehdreepilzniocneauislt)aheearnodtchcceloumadppeinoctsaelesdowroefrbea2s1do.el8aet(pe4r(.a1l5.)4plfiabsr0im.la7s CanadT+a+bclh,elaonIrdisdFheeow+ws+ertiehnedtaehtneeacldtyaozbmelidencaaonmntosutniittousnes.,ntTshwieotfohAsmGKoFl+.a,liSMtoygdiw+ua+ms,
(Fig. 6). Most of the fibrils were parallel to the apical near that ofplasma (750-875 mOsm), and significant con-
plasma membrane, with those constituting the basal one- centrations of protein and urea were measured. The pH
fourth of the zonulae occludentes forming a loose anasto- varied between 8.0 and 8.2.
mosing network (Fig. 6).
Ultrastructural observations of the solids from alkaline Electrophysiology
gland fluid showed cellular debris including multivesicular Baseline parameters. The baseline PD was 14.5 1.9
bsotdaiineesd, psopshietriivceally pfaorrtilcilpeosfuwsicitnh, emleemctbrroann-deenwsheorclosr,esantdhaat tmaVn,ceIswcawsasca2l9c.u1lated4.t2ojbueA/c7m322,.4and tr1an8s4e.p6ithoehlimal recsinsr-
few necrotic spermatozoa (Fig. 7). =
(n 18).
Transport inhibitors. The effect of specific ion transport
Localization ofcarbonic anhydrase activity inhibitors on the baseline Isc is shown in Table II. The
Carbonic anhydrase activity (CAH) was indicated by a serosal addition of ouabain, a Na+/K+ ATPase inhibitor,
black precipitate at both the light and electron microscopic resulted in an almost 48% decrease of Isc within 45 to 50
level (Figs. 8, 9). A minimum of 2 min in the incubation min. Bumetanide, a Na+/K+/CF cotransport inhibitor, de-
medium was required for the precipitate to develop, at creased the Isc approximately 70% within 30 min, and
which time CAH appeared first within the intercellular DIDS, a Cr/HCO3-exchange inhibitor, placed on the lumi-
space ofcolumnarcells. In electron micrographs, CAH was nal side of the epithelium decreased the Isc almost 38%
localized in the intercellular space between columnar cells within 30 to 40 min. The Isc was completely inhibited, and
but excluded from the zonulae occludentes (Fig. 9). Adja- in fact was slightly reversed, after consecutive addition of
cent to the basement membrane, CAH was observed only bumetanide within 30 min ofDIDS addition to the lumenal
within the intercellularspace formedby invaginations ofthe surface. The effect ofluminal exposure to acetazolamide, a
plasma membrane or interdigitation of cytoplasmic folds carbonic anhydrase inhibitor, on Isc was sporadic, and pro-
(Fig. 9). Regions of the basolateral plasma membrane that duced only a 16% overall reduction of Isc. Amiloride, a
contacted the basement membrane did not exhibit CAH. sodium ion channel inhibitor, placed in eitherthe luminal or
After3-10minofincubation, theprecipitate appeared inthe serosal media had no significant effect on the Isc (data not
basal two-thirds of columnar cell cytoplasm (Fig. 8). shown). The removal ofchloride from the bathing media on
Control sections incubatedonbicarbonate-free Hansson's both sides of the tissue with the substitution of gluconate
medium or on ammonium sulfide alone were similar to resulted in a 67% reduction in Isc by 45 min.
unstained sections that were rinsed only on Sorensen's Alkalinization rates. Two experiments investigating de-
Figure 1. Scanning electron micrograph ofatransected mucosal fold. The asterisk is located in thecenter
of an arteriole. A network of capillaries (arrows) lies directly beneath the epithelium, which has prominent
secondary lysosomes (arrowheads). Bar = 50 /im.
Figure2. Transmissionelectronmicrograph(TEM)ofthesimplecolumnarepitheliumofthealkalinegland.
Arrows indicate secondary lysosomes located in the supranuclear region. Arrowheads indicate a nerve fiber
closely adjacent to the epithelium. Note the numerous vesicles in the apical cytoplasm. Bar = 2 fj.m.
Figure3. Light micrograph (LM)ofnerve fibers (arrows) in the subepithelial laminapropnastained black
usingtheSevierMungersilvertechnique.Nerveliberswerecloselyassociatedwilhbloodvessels(asterisks)and
the epithelium (e). Note the multiple darkly staining secondary lysosomes in the apical cytoplasm of the
epithelium. Bar = 4 /j.m.
Figure4. TEMofabasalcell(BOlocatedbetweenadjacentcolumnarcells(CC).Notethelargeproportion
ofthe nucleus relative to the BC cytoplasm. Basal lamina (BL). Bar = 2 /tim.
G. M. GRABOWSKI ET AL
Figure 5. Figure 7.
STINGRAY ALKALINE GLAND
Table I Table II
Analyzedconstituents ofalkaline glandfluid Effects ofion transport inhibitors on the short-circuit current(Isc)
Component
90 G. M. GRABOWSKI ET AL
Table III
In \ilroalkalini-alion mles cj'alkiilinc ^ln
Alkalinization rate
Experimental conditions ofacid/cnr/h)
Experiment I
STINGRAY ALKALINE GLAND 91
Freeze fracture replicas showed that the only distinguishing transporter at the basolateral plasma membrane, and a chin-
intramembranous particles were in the basolateral plasma ride channel at the apical plasma membrane.
membrane. The size and distribution of the particles form- Chloride movement across the epithelial basolateral
ing these clusters was comparable to those forming gap plasma membrane, via a putative Na+/K+/CF cotrans-
junctions in mammalian cells. Apical and basolateral porter in epithelial cells in stingray alkaline gland, appears
plasma membranes did not reveal any rod-shaped particles to be driven in part by Na^/K+ ATPase, as shown by the
that would suggest proton transport (Brown and Montesano, serosal addition of ouabain, which decreased the Isc by
48%. In contrast, ouabain completely abolished chloride
1980).
The zonulae occludentes of columnar cells consisted of secretion and Isc in the little skate alkaline gland (Smith.
about 22 strands, suggesting that the epithelium is electri- 1985).
cally tight and imparts a high transepithelial resistance The lack ofsignificant alkalinization rates after the tissue
(ClaudeandGoodenough. 1973; Claude, 1978). Ourin vitro was exposed to medium free of chloride and sodium sug-
electrophysiological data showed that the transepithelial gests thatthere is little independenttransportofbicarbonate.
mreosripshtoanlcoegyw.asTh73e2porehsmenccemo2f, c"ovnefriyrmtiingght"thezotniughltaejunocctciloun- aIfnaaspiigcnailfiCclant/pHoCrtOi,oneoxfcthhaenaglekralinaiszaotuironrpesruolctessssuigngveosltves
dentes and a high transepithelial resistance suggests that the absence of luminal chloride could impede that process,
there is little paracellular solute transport across the epithe- resulting in the accumulation of intracellular bicarbonate.
lium ofthe alkaline gland (Bowman el al, 1992; Byers and Such a scenario has been observed in the rat parotid acini:
Marc-Pelletier. 1992). Therefore, regulation ofion transport SITS, an inhibitor ofbicarbonate transport, increased intra-
appMeaarresntoetocalc.ur(1p9r6i3m)arialnydaScmriotssh t(h1e98p1l.as1m9a85m)emhbarveandee.m- Ttcieuolrnlnuelvrai.rap1aH9n9i2ao)n.ndCcwhhalasonrntiehdloesugc(hhPtainrtnaoneilstsietmiunallaa.t.enu1bm9i8bc7ae;rrboMonefaltdveiifnsfeecrarenend-t
onstrated that both bicarbonate and chloride are secreted in
athneiolnittrleesspkoantseiballekalfionremgolsatndoafndthethaItscc.hlTohriisdefiisndtihnegmwaains erbpeoisntpahiterleaita(o.Gryrianeycplieuttdhiealln.ig.a,p1a9hn8ac9vr;eeaTtbaiebcecnhdaucrstah,noiswwneetatatot..gtlr1aa9nn8sd9p;odrutKcutbn,izcaealnr-d-
einxtwenhdiecdhtIoscthedesctriengarsaeydaalklamloisnte g7l0a%ndwinhethne pcrhelsoernitdestwuadsy mann et aI.. 1991 ) at a conductance as high as 50% of the
conductance of chloride.
removed from the bathing medium. Using intracellular mi- The remaining Isc may be accountedforby aNa+/HCOJ
croelectrodes. Smith (1981, 1985) showed that the apical symport, as demonstrated in this study by using pH stat
plasma membrane was dominated by a large chloride con- methodology. Such mechanisms for bicarbonate transport
ductance, whereas the basolateral plasma membrane con- have been demonstrated in renal proximal tubule (Yoshi-
tained a barium-sensitive potassium channel. However, the tomi et til.. 1985), corneal endothelial cells (Wiederholt et
mechanisms involved in the alkalinization process have ill.. 1985), and gastric oxyntic cells (Curci et al.. 1987).
never been clearly established in this gland, despite specu- Alkalinization ofthe luminal medium in the present study
lationthataCr/HCO3 exchangermayexist inthe apical or was dependent on the presence of both apical chloride and
basolateral plasma membrane orin both membranes (Maren serosal sodium. The changes attributed to the absence and
et al.. 1963; Smith. 1981. 1985). readdition ofsodium suggests the presence ofaNa+/HCO^
serIonsatlheaddpirteisoenntofsbtuudmye,tatnhiedem.arakneidnhirbeidtuocrtioofnNain+/IsKc+/afCtle~r osfympseorrots.alThseodailukma,linainzdatiiotsn rraetdeucattitornibubtyedltuomitnhealreSaIdTdSit,ioins
cotransport. suggests that this transporter is located in the indirectevidence that aNa4/HCO^ symport maybe located
basolateral plasma membrane. If so. it may be the main at the apical plasma membrane. The stilbene. SITS, blocks
conductive pathway for chloride entry into the cell. The not only bicarbonate transport via Na'/HCO, symporters
remaining Isc could be due to the secretion of intracellular (Curci et al.. 1987; Fitz et al., 1989; Wiederholt et al..
chloride or another anion. such as bicarbonate. To test this 1985), but also Cl /HCO, exchangers (Stewart et al..
latter possibility we added the stilbene, DIDS, which effec- 1989).
tively inhibits bicarbonate cotransporters (Wiederholt ettil.. Maren and co-workers (1963) demonstrated a possible
1985; Melvin and Turner. 1992) as well as chloride chan- relationship between CAH and higher pH levels in AGF of
nels (Bretag. 1987) to the luminal side of the epithelium. various skate species. They showed that inhibition ofCAH
The resultant 38% decrease ofIsc. and its further reduction /;; vivoreduced the pH ofnewly formed fluidtolevels found
to nominal levels after the consecutive addition of serosal in species that did not have glandular CAH. This was
bumetanide. substantiates this assumption. Furthermore, accomplished using intravenous injections ofacetazolamide
complete reduction of the Isc by consecutive addition of at least 10 times higher than the dose we used. In a study of
DIDS and bumetanide suggests a pathway for chloride rat distal colon, the need for high (millimolar) concentra-
secretion across the epithelial cells via a Na+/K+/CP co- tions of acetazolamide to inhibit bicarbonate transport was